Hi Chris, thanks for your reply
The script is still functionally correct and can be executed successfully. Its more of a source control traceability issue. Unfortunately we do not have personal databases dedicated to a certain project or piece of work, so when we produce schema comparisons, we often only want to include a subset of the changes in the deployment script. Being able to then compare the deployment script with a previous deployment to ensure we have only captured the changes we want is important. It makes it difficult when objects appear in different parts of the script and are identified as differences by text comparison tools.
I assume there is a degree of ordering implemented to ensure dependencies are applied in the correct order etc. Apart from that, the ordering seems arbitrary as sometimes it is the same as previously, but not always. I thought it might be simple enough to add additional ordering criteria to ensure a specific order is retained each time.